NOMMU and FRV patches

Hi Linus,

Can you pull these patches please?

There are a number of fixes for NOMMU: fixes for percpu to deal with breakage due to MMU-mode changes and fixes for the firmware loader to make it work at all.

There are a number of fixes for FRV: a missing argument and a set of IRQ patches from Thomas Gleixner.

David
---
The following changes since commit 4bbba111d94781d34081c37856bbc5eb33f6c72a:

Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6 (2011-03-23 07:58:09 -0700)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-2.6-frv.git frv

Amerigo Wang (1):
FRV: Do some cleanups

David Howells (2):
NOMMU: percpu should use is_vmalloc_addr().
FRV: Missing node arg in alloc_thread_info_node() macro

Graf Yang (1):
NOMMU: support SMP dynamic percpu_alloc

Mike Frysinger (2):
NOMMU: new vcoalesce helper function
firmware loader: use vcoalesce for NOMMU

Thomas Gleixner (7):
frv: Remove stale irq_chip.end
frv: Fix typo from __do_IRQ overhaul
frv: Convert mb93091 irq_chip to new functions
frv: Convert mb93093 irq_chip to new function
frv: Convert mb93493 irq_chip to new functions
frv: Convert cpu irq_chip to new functions
frv: Select GENERIC_HARDIRQS_NO_DEPRECATED

 arch/frv/Kconfig                   |    2 +-
 arch/frv/include/asm/system.h      |    9 ---------
 arch/frv/include/asm/thread_info.h |    4 +++-
 arch/frv/kernel/irq-mb93091.c      |   26 +++++++++++++-------------
 arch/frv/kernel/irq-mb93093.c      |   29 ++++++++++++++---------------
 arch/frv/kernel/irq-mb93493.c      |   22 +++++++++++-----------
 arch/frv/kernel/irq.c              |   35 +++++++++++++++--------------------
 drivers/base/firmware_class.c      |    2 +-
 include/linux/vmalloc.h            |   35 +++++++++++++++++++++++++++++++++++
 mm/nommu.c                         |   22 +++++++++++++++++++++-
 mm/percpu.c                        |    3 +--
 11 files changed, 115 insertions(+), 74 deletions(-)

No.

Because they are sh*t that have clearly never been tested, and have never been in -next.

They don't even compile on anything but NOMMU, as far as I can tell. That new "vcoalesce()" helper is crap:

drivers/base/firmware_class.c: In function ‘firmware_loading_store’:
drivers/base/firmware_class.c:251:29: error: expected expression
before ‘...’ token
drivers/base/firmware_class.c:251:29: error: too few arguments to
function ‘vmap’
include/linux/vmalloc.h:70:14: note: declared here

please send it to me next merge-window, when it has actually sat in -next.

Linus

yeah, it would seem that my patch broke MMU (for once the other way around). it's the MMU vcoalesce redirection that is broken. stupid thinko:
-#define vcoalesce(...) vmap(...)
+#define vcoalesce(...) vmap(__VA_ARGS__)
-mike